Identificador persistente para citar o vincular este elemento: http://hdl.handle.net/10553/47342
Título: Gains tuning of a PI-Fuzzy controller by genetic algorithms
Autores/as: Betancor-Martin, Carlos S. 
Sosa González, Carlos Javier 
Montiel-Nelson, Juan A. 
Vega-Martinez, Aurelio 
Clasificación UNESCO: 3307 Tecnología electrónica
Palabras clave: Robust control
Genetic algorithms
Optical control
PID control
Artificial neural network, et al.
Fecha de publicación: 2014
Editor/a: 0264-4401
Publicación seriada: Engineering Computations 
Resumen: Purpose - Nowadays, in order to improve current applications, industry incorporates to their solution approaches artificial intelligence techniques and methodologies like Fuzzy Logic, neural networks and/or genetic algorithms (GA). Artificial intelligence techniques complement classical methodologies and include concepts that simulate the way humans solve problems or how processes work in nature. In this work, the Fuzzy Logic system cancels the effects of load perturbances in an energy plant, by implementing a secondary controller which complements the main controller. The purpose of this paper is to use GA to tune this new secondary controller. The authors particularize the proposal for three specific applications: control the angular speed and position of a Direct Current (DC) motor and control the output voltage of a DC/DC buck converter.Design/methodology/approach - The authors use GA for tuning a Proportional-Integral Fuzzy Controller (PI-Fuzzy). The proposal defines a new objective function in comparison with literature approaches. The main key in the new objective function is combining the best features of Integral Square Error (ISE) function and taking out the overshoot response.Findings - In order to demonstrate the proposed methodology based on GA tuning a PI-Fuzzy, the authors apply the literature benchmark to the solution. The results are compared with the following techniques: Robust control, continuous PID control, discrete PID control, Optimal Control, Fuzzy Control and Artificial Neural Network based control. Comparisons are presented in terms of setting time and overshot.Originality/value - Results demonstrate that ISE or integral of absolute value of error function do not provide the desired response. Achieved results demonstrate the usefulness of the proposal to eliminate the overshoot of the traditional behaviour without lost any of the main features of the literature methodologies.
URI: http://hdl.handle.net/10553/47342
ISSN: 0264-4401
DOI: 10.1108/EC-03-2012-0068
Fuente: Engineering Computations (Swansea, Wales)[ISSN 0264-4401],v. 31, p. 1074-1097
Colección:Artículos
Vista completa

Google ScholarTM

Verifica

Altmetric


Comparte



Exporta metadatos



Los elementos en ULPGC accedaCRIS están protegidos por derechos de autor con todos los derechos reservados, a menos que se indique lo contrario.